Blade server architecture is an evolving technology that conveniently packages a server on a single board called a processing blade and houses a plurality of similar processing blades in a chassis. The processing blades can be easily installed or removed as desired and share common resources, such as input/output (“I/O” ports), a compact disc read only memory (“CD-ROM”) drive, a floppy drive, a digital versatile disc (“DVD”) drive, one or more network connections, etc. The ease with which the processing power can be scaled by adding or removing processing blades is a key feature driving the development and use of blade server technology. This scalable feature makes blade servers ideal candidates for performing tasks such as data and webpage hosting.
Processing blades typically are complete processing systems including one or more processors, firmware, system memory, one or more hard disks, one or more network interface cards (“NICs”), and the like. Known processing blades also generally include a baseboard management controller (“BMC”) and an associated BMC firmware unit. A BMC is usually a coprocessor or service processor, separate from the one or more main processors. The BMC communicates with a chassis management module (“CMM”), typically mounted on the chassis of the blade server, to coordinate and to manage operation of the individual processing blade.
In general, a blade server will house a plurality of processing blades, each having its own BMC and BMC firmware unit, which all coordinate with the CMM. The inclusion of the BMC and the BMC firmware unit in each processing blade substantially increases build of material (“BOM”) costs. In addition, the BMCs and the BMC firmware units consume valuable real estate on the surface of the processing blades. These two adverse consequences of the BMCs and the BMC firmware units directly conflict with two key goals associated with the development and implementation of blade server technology—providing a high density and economical processing solution.